KR20110119671A - Safety system for a motor vehicle - Google Patents

Abstract

A safety system for a motor vehicle comprises a sensing arrangement (11) providing sensor signals related to the surrounding environment of the vehicle, at least one safety means (13, 14, 15, ...) for an occupant of the vehicle, and a control means (22) adapted to control said safety means (13, 14, 15, ...) depending on signals from said sensing arrangement (11). The safety system (10) comprises an environment classifying means (23) adapted to classify the surrounding environment of the vehicle into different predetermined categories on the basis of signals from said sensing arrangement (11), and to adjust said control means (22) depending on the vehicle environment category determined by said environment classifying means (23).

Description

Safety system for a motor vehicle

The present invention relates to a safety system for a motor vehicle, comprising: a sensing device for providing sensor signals to a surrounding environment of the vehicle, at least one safety means for an occupant of the vehicle, and the safety means in accordance with signals from the sensing device. Control means adapted to control the control.

In safety systems for motor vehicles, means are known for detecting objects in front of the vehicle, for example pedestrians, cyclists and / or animals. However, such detection means may still be susceptible to false detection, which irritates the driver, more importantly the user distrusting the safety system and, more seriously, the system fails to alert the driver by missing the detection. Done. Moreover, such detection means typically consume a lot of processing capacities, which is one of the reasons why detection reliability is limited.

It is an object of the present invention to provide a safety system for a motor vehicle with high detection reliability and low probability of false detection.

The present invention solves the above object as a feature of claim 1. Categorizing the surrounding vehicle environment into different categories allows adjusting the safety system according to the determined vehicle environment. In particular, the algorithm for detecting the object in front of the vehicle can be optimally selected or adjusted according to the detected vehicle environment, which leads to increased detection reliability and error-free detection. Moreover, the present invention allows deactivation of algorithms in control means deemed irrelevant to the determined vehicle environment, contributing to the effective use of processing resources and thereby leading to improved detection reliability.

 Various predetermined categories of vehicle environments may include urban environments, nonurban environments, rural environments, highway environments, and the like. These categories may also be subcategories. Preferably the environmental classification means is adapted to classify the environment of the vehicle into at least an urban environment and a non-urban environment.

Preferably the safety measures are adapted to provide safety related information to the vehicle occupant. A particularly advantageous application of the present invention involves providing a warning to the vehicle occupant if the animal is detected in front of the vehicle. In such an application, environmental classification means are preferably adapted to determine the urban environment of the vehicle. If the urban environment is judged by environmental classification means, the animal detection means of the safety system is preferably deactivated, since the probability of vehicle-animal collision is assumed to be zero in the urban environment. As a result, detection of any false wildlife in an urban environment is prevented. Moreover, processing resources saved due to deactivation of animal detection can be allocated to other processes, for example to pedestrian detection, whereby the reliability of pedestrian detection can be further improved thereby. .

Moreover, environmental classification means are preferably adapted to determine the rural or non-city environment of the vehicle. If the rural or non-urban environment has been determined by the environmental classification means, the animal detection means of the safety system is preferably activated.

Other predetermined states result in animal warning deactivation, and activation may be selected in the control means. In general, the control means can be adjusted for one of a plurality of predetermined states in accordance with the determined vehicle environment.

In a preferred embodiment, the environmental category is determined based on the image data provided by the imaging means of the sensing device. The imaging means may comprise one or a plurality of cameras facing the area in front of the vehicle, in particular an optical camera or an infrared camera. The vehicle environment category may be determined from the detected image by using an image analysis technique, and may be determined using, for example, a pattern recognition technique in which a principle is known in the art. Other methods may be employed to distinguish between predetermined vehicle environment categories.

The environmental category may be determined based on vehicle data provided by sensors mounted on one or more vehicles, for example based on vehicle speed and / or yaw rate. For example, to determine the urban environment, the fact that the probability of rotation in the urban environment is higher and / or that the vehicle is slower on average movement can be used. Thus, it is possible to use known time and / or speed ratios from past time window classifications that classify the vehicle environment into urban or non-urban environments.

In a particularly preferred embodiment, the combination of the image data provided by the image means and the vehicle data provided by at least one of the vehicle sensors is used to determine the environmental category. For example, image data can be used to determine the environmental category and vehicle data can be used to identify the detected environmental category, which improves the reliability of the detected vehicle environment category.

According to the present invention, the vehicle environment may be classified and used in various categories, and detection and warning of pedestrians and animals may be enabled.

In the following, the invention will be described based on preferred embodiments with reference to the attached drawings.
1 is a schematic view of a safety system.
2 is a process diagram of an animal warning system in a safety system.

The safety system 10 is mounted on a motor vehicle and comprises a sensing device 11, an electronic processing means 12 and a plurality of safety means 13, 14, 15 .., for the vehicle occupant, for example a spare Pre-tensioner means 13, airbag means 14, animal warning means 15 and the like.

The sensing device 11 preferably comprises an image means 20 and a vehicle sensor 18, 19, wherein the image means is configured to record an image of the front area of the vehicle, the vehicle sensor comprising data relating to the movement of the vehicle. And a yaw sensor 18, a speed sensor 19 and other vehicle mounted sensor means. The sensing device 11 may include other sensors such as a radar sensor device. The imaging means 20 can be optical and / or infrared imaging means, the infrared rays encompassing near IR and / or far IR. Preferably the imaging means 20 comprise a camera device, here in the form of stereo cameras 16 and 17.

The programmable electronic processing means 12 preferably comprise a microprocessor or microcontroller and can be implemented in an electronic control unit (ECU), for example mounted in a vehicle. The sensor device 11 and the safety means 13, 14, 15... Are connected to the electronic processing means 12 via the on-vehicle data bus 21.

The processing means 12 comprise control means 22 adapted to process the signals received from the sensor device 11 and to control the safety means 13, 14, 15... According to the processing result. . For example, the pre-tension means 13 and / or airbag means 14 when the pre-shock situation and / or the shock situation have been determined based on signals from the corresponding (pre) shock sensors of the sensor device 11. Can be activated by the control means 22. The control means 22 are preferably implemented by software.

The processing means 12 further comprises an environment classification means 23, preferably it is implemented by software, the environment classification means processing the signals received from the sensor device 11, as indicated by the arrows in FIG. 1. It is adapted to adjust the control means 22 according to the same processing result. In particular, the environmental classification means 23 is adapted to determine the category or class of the surrounding environment in which the vehicle is running. More specifically, a plurality of models or one model for a predetermined plurality of possible environment categories EC1, EC2 ... ECn are stored in advance in the environment classification means 23. The environment classification means 23 comprises an environment classifier, which measures a closeness measure between each predetermined environment category (EC1, EC2 ... ECn) and the current environment-related data 25. Processing the environment-related data 25 received from the sensor device 11 to classify the current vehicle environment into one of the predetermined environment categories EC1, EC2 ... ECn. For example, the environment classifier 24 may calculate, for the current set of environment-related data 25, the probability that the current vehicle environment will belong to each of the predetermined environment categories EC1, EC2... ECn. Next, the predetermined environment category ECi having the highest probability value is determined to be the current environment category. To improve the reliability of the classification, the environment classifier 24 is configured to perform classification filtering in real time over time by considering the weighted sum of several time consecutive classificatoin results. Can be adapted.

The environment classifying means 23 further comprises a determining means 26, which determines different control signals for setting various predetermined states 27, 28 ... in the control means according to the determined environmental category. Adapted to provide to the means 22. In this way the control software in the control means 22 for controlling the safety means 13, 14, 15... Can be optimally adjusted for the current vehicle environment.

The environmental categories EC1, EC2... ECn may include, for example, urban environment, non-urban environment, highway environment, rural environment and / or other environmental categories. In order to allow the determination of the environmental category, the input data 25 for the environmental classification means 23 must be directly or indirectly related to the environment of the vehicle. In one embodiment the data from the sensor of the sensor device 11 which directly senses the area in front of the motor vehicle is used as input data 25 to the environmental classification means 23. In particular it may be data from the camera system 20 or the radar sensor device.

In another embodiment, vehicle motion data from vehicle motion sensors 18, 19, such as yaw sensor 18 and / or speed sensor 19, may be used as input data 25 for environmental classification means 23. Can be. The driving data of such a vehicle is directly related to the environment of the vehicle because, for example, in an urban environment, there is a high probability that the vehicle will always move slower and rotate as compared to a non-urban environment.

In a preferred embodiment the data from the sensor 20 directly sensing the area in front of the vehicle and the data from the vehicle motion sensors 18, 19... Are input data 25 for the environmental classification means 23. ), Which contributes to higher classification reliability.

Next, a preferred example of applying the above-described features to the animal warning system is described. In this application the control means 22 are adapted to detect an animal in the area in front of the motor vehicle based on the image data provided by the image means 20, such animals being elk, deer, horse, As a large animal such as lights out, there is a great danger in collision with a vehicle. Technically this can be implemented on the same principle as pedestrian detection from image data.

When the control means determines the animal in the area in front of the vehicle which is potentially at risk of colliding with the vehicle, the control means 22 activates the animal warning means 15 which provides the animal warning to the driver. Animal warnings can be in any suitable form and include warning, visual and / or tactile warning signals, for example flashing monitors, warning sounds and / or warning information displayed on the driver's display. to be.

In a particularly preferred embodiment, the environmental classification means 23 is adapted to classify the current environment into two environmental categories, i.e. to classify into an urban environment and a non-urban environment. When the urban environment is determined by the classifier 24, the judging means 26 sets the control means 23 to a state 27 in which the animal detection and the animal warning are deactivated. This is based on the assumption that large animals do not appear in urban environments. Processing resources associated with animal detection can be assigned to other processes, such as pedestrian detection, which improves the reliability of the safety system.

On the other hand, in the case where the danger of the non-urban environment and thus the large animal is increased by the classifier 24, the decision means 26 is the control means 28 in which the animal detection and the animal warning are activated. Set (22). In this case, processing resources can be shared between pedestrian detection and animal detection.

As mentioned above, two or more environmental categories may be determined by the environmental classification means 23, and / or the environmental classification means 23 may set two or more states to the control means 22. For example, if a freeway environment is determined, pedestrian judgment and warning can be deactivated and only animal detection and warning can be active because it can be assumed that no pedestrian is present in the freeway environment.

The processor of environmental classification and control adjustment of the safety means based on the determined environment is permanently repeated in real time during the operation of the vehicle, and at a frequency suitable for detecting changes in the vehicle environment in time.

The invention is not limited to adjusting the control of the animal warning means in accordance with the determined vehicle environment. The present invention includes adjusting the control of any other vehicle occupant safety means, such as, for example, preliminary tensioners and / or airbag means, depending on the determined vehicle environment. The term safety means generally encompasses occupant protection measures as well as occupant warning measures.

11.Sensing device 13.14.15. Safety measures
22. Control means 23. Classification means

Claims (15)

A sensing device 11 providing sensor signals for the surrounding environment of the vehicle, at least one safety means 13, 14, 15... For the occupant of the vehicle, and signals from the sensing device 11. As a safety system for a vehicle 10, comprising control means 22 adapted to control the safety means 13, 14, 15...
The safety system 10 has environmental classification means 23 adapted to classify the surrounding environment of the vehicle into different predetermined categories based on the signals from the sensing device 11, and also the environment. The classification means (23) is characterized in that it is adapted to adjust the control means (22) according to the vehicle environment category determined by the environment classification means. The method of claim 1,
The safety means (13, 14, 15 ...) are adapted to provide safety related information to the vehicle occupant. The method according to claim 1 or 2,
The safety system (10) is adapted to detect an animal in front of the vehicle. The method of claim 3, wherein
The safety system (15) is adapted to provide the vehicle occupant with an animal warning as to whether or not the animal has been detected in front of the vehicle. The method of claim 1, wherein
The environment classification means (23) is adapted to determine whether the environment surrounding the vehicle is an urban environment. The method of claim 5, wherein
The animal detection means of the safety system (10) is deactivated if the urban environment is determined by the environmental classification means (23). The method of claim 1, wherein
The environmental classification means (23) is adapted to determine that the environment surrounding the vehicle is a rural environment or a non-city environment. The method of claim 7, wherein
The animal detection means of the safety system (10) is activated if a rural or non-urban environment has been determined by the environmental classification means (23). The method of claim 1, wherein
The environmental classification means (23) is adapted to perform classification filtering over time by considering the weighted sum of several time consecutive classification results. The method of claim 1, wherein
The control means (22) can be adjusted for one of a plurality of predetermined states in accordance with the determined vehicle environment. The method of claim 1, wherein
The environmental category is determined on the basis of the image data provided by the image means (20) of the sensing device (11). The method of claim 1, wherein
The environmental category is determined based on vehicle data provided by at least one vehicle sensor (18, 19) of the sensing device (11). The method of claim 1, wherein
The environmental category is determined on the basis of the speed data provided by the speed sensor (19) of the sensing device (11). The method of claim 1, wherein
The environmental category is determined based on yaw rate data provided by a yaw sensor 18 of the sensing device. The method of claim 1, wherein
The environmental category is determined based on a combination of the image data provided by the image means (20) and the vehicle data provided by the at least one vehicle sensor (18, 19).

Images